|
|
Post by qdtjni on May 4, 2016 18:32:34 GMT -5
|
|
|
|
Post by yves on May 4, 2016 19:30:27 GMT -5
Face the fact low amounts of supersonic noise are still undetectable by humans?  Are you stating that your DACs specifications of SNR are for a non limited frequency range? Limited in some way of course yes, but certainly not limited to 20 kHz. How else do you imagine it would be at all possible for the Dynamic Range of my DAC to, in spite of it having "only" 98 dB SNR, still arrive at a very respectable 130 dB? Not a pissing contest, unless perhaps the primary reason for your continued postings in this thread is to turn it into exactly that? There are people who want to drive their digital data to their DAC in the trunk of a Cadillac or Rolls Royce, and who will fantasize about that kind of stuff, whereas others feel the need to focus more on the final quality outcome in the analog output of the DAC. |
|
|
|
Post by yves on May 4, 2016 19:50:34 GMT -5
Nice try. In 2007, there were no cost-effective, high-performance asynchronous USB 2.0 interface implementations to compare, as they simply had not been developed yet. |
|
|
|
Post by qdtjni on May 5, 2016 3:58:03 GMT -5
Nice try. In 2007, there were no cost-effective, high-performance asynchronous USB 2.0 interface implementations to compare, as they simply had not been developed yet. It is a player with the DAC combined using i2S having from 2007 way better SNR than your contemporary DAC.. |
|
|
|
Post by qdtjni on May 5, 2016 4:02:39 GMT -5
Are you stating that your DACs specifications of SNR are for a non limited frequency range? Limited in some way of course yes, but certainly not limited to 20 kHz. How else do you imagine it would be at all possible for the Dynamic Range of my DAC to, in spite of it having "only" 98 dB SNR, still arrive at a very respectable 130 dB? Not a pissing contest, unless perhaps the primary reason for your continued postings in this thread is to turn it into exactly that? There are people who want to drive their digital data to their DAC in the trunk of a Cadillac or Rolls Royce, and who will fantasize about that kind of stuff, whereas others feel the need to focus more on the final quality outcome in the analog output of the DAC. The question you should ask is why does it have that SNR when the Dynamic Range is so good. If the SNR specified for your DAC is for wideband, it's really not bad at all. Would be good to see if the given value is A weighted or Wideband. You can compare it with the Mytek Manhattan using the same ESS DACs; everythingaudionetwork.blogspot.ch/2015/10/ean-speccheck-benchtest-measurement.html |
|
|
|
Post by yves on May 5, 2016 5:54:58 GMT -5
Nice try. In 2007, there were no cost-effective, high-performance asynchronous USB 2.0 interface implementations to compare, as they simply had not been developed yet. It is a player with the DAC combined using i2S having from 2007 way better SNR than your contemporary DAC.. The notion that a better wideband SNR is still key to advancing fidelity comes from an outdated concept of resolution. |
|
|
|
Post by yves on May 5, 2016 6:35:31 GMT -5
Limited in some way of course yes, but certainly not limited to 20 kHz. How else do you imagine it would be at all possible for the Dynamic Range of my DAC to, in spite of it having "only" 98 dB SNR, still arrive at a very respectable 130 dB? Not a pissing contest, unless perhaps the primary reason for your continued postings in this thread is to turn it into exactly that? There are people who want to drive their digital data to their DAC in the trunk of a Cadillac or Rolls Royce, and who will fantasize about that kind of stuff, whereas others feel the need to focus more on the final quality outcome in the analog output of the DAC. The question you should ask is why does it have that SNR when the Dynamic Range is so good. If the SNR specified for your DAC is for wideband, it's really not bad at all. Would be good to see if the given value is A weighted or Wideband. You can compare it with the Mytek Manhattan using the same ESS DACs; everythingaudionetwork.blogspot.ch/2015/10/ean-speccheck-benchtest-measurement.html Obviously the 98 dB SNR measurement is wideband. Interestingly, the $4,995 Mytek does not yield higher Dynamic Range than my $1,350 Eastern Electric; the latter brings on-the-fly switching between solid state gain stage and tube based gain stage to the table, and can be upgraded (optionally) with femto clocks and discrete op-amps because the design is modular of course so the buyer gets the additional freedom of choice on whether to break the bank towards scraping that last grain of sand from under the soles of the pair of ES9018 chips or towards optimum balance of maybe drinking more excellent beer instead. I like that kind of freedom a lot.  |
|
|
|
Post by qdtjni on May 5, 2016 6:49:54 GMT -5
It is a player with the DAC combined using i2S having from 2007 way better SNR than your contemporary DAC.. The notion that a better wideband SNR is still key to advancing fidelity comes from an outdated concept of resolution. While a low widened SNR not necessarily say anything about the more important SNR limited to hearing range or A weighted, a high wide band SNR implies a high a SNR limited to the hearing range as well as A weighted SNR. It can also mean that a DAC has a good filter design. |
|
|
|
Post by qdtjni on May 5, 2016 7:17:54 GMT -5
Obviously the 98 dB SNR measurement is wideband. Interestingly, the $4,995 Mytek does not yield higher Dynamic Range than my $1,350 Eastern Electric; the latter brings on-the-fly switching between solid state gain stage and tube based gain stage to the table, and can be upgraded (optionally) with femto clocks and discrete op-amps because the design is modular of course so the buyer gets the additional freedom of choice on whether to break the bank towards scraping that last grain of sand from under the soles of the pair of ES9018 chips or towards optimum balance of maybe drinking more excellent beer instead. I like that kind of freedom a lot. There's nothing obvious about not stating whether a measurement is, wide band, unweighted or A weighted, on the contrary. As for having the same dynamic range, that's to be expected when they share the same DAC chip setup. DUH! Anyway, the point I was trying to make is that the SNR for your DAC is pretty good, given that it really is the wide band value. Comparing prices for two so fundamental different products does not make sense, it's very much apples to oranges. Apart from sharing the basic DAC through USB, S/PDIF and TOSLINK, the Mytek Manhattan has way more functionality, where your DAC offers switchable gain stage. Different horses for different courses. As for your personal finances, not really interested.  |
|
KeithL
Administrator 
Posts: 10,274
|
Post by KeithL on May 5, 2016 9:53:02 GMT -5
Methinks y'all need to come on up out of that there rabbit hole.... DACs are really a somewhat separate topic than what this started out as... we've gone from comparing horses to automobiles to comparing the exact color tint on different types of headlights. As per the debate between "separate DAC" and "all-in-one unit", the realities are that anything you can do in a separate DAC you can do in the DAC in an all-in-one unit. By putting it all in one box, you reduce the cost, and the footprint, but you make it more likely that the computer stuff will interfere with the DAC stuff, which means more careful design, which will probably push the cost back up again for comparable performance. Also, since the two are in fact tied together, you can't choose, or upgrade, them separately. It's pretty well all the same arguments used to justify whether to use a separate preamp and power amp - or go with an integrated unit - except that computer stuff is prone to very frequent updates. And, yeah, DAC guys tend not to like to design computers, and computer guys don't like to design audiophile DACs. Let me clarify or comment on a few other points here as well...... 1) Dynamic range is intimately related to SNR. While it's technically possible to "discern" information that's below the noise floor, which makes for a slight technical difference between them, for most practical purposes you can consider dynamic range and SNR to be one and the same. (DR is simply the difference between the loudest sound and the quietest sound that a device can deliver - expressed as a dB ratio. In general, anything that's much quieter than the noise floor isn't going to be very useful. This is doubly true if the noise floor is low enough to be inaudible to begin with.) 2) There are several different ways of specifying things. For example, if I specify S/N ratio, I'm telling you the ratio between SIGNAL and NOISE. But, if I talk about THD+N, then I'm clumping the distortion and noise together in the measurement. (This makes sense because I'm simply comparing "signal" to "all the other junk that doesn't belong there". With an amplifier, THD+N will often be much higher than noise alone; since DACs usually have very low THD, this usually isn't the case.) A-weighting is simply a way to specify S/N that takes into account the fact that certain frequencies are more audible than others - so EQ is applied before you calculate the measurement. The A-weighted measurement will always be better than the unweighted measurement, but the two can't be mathematically related - the size of the difference will depend on the spectrum of the specific noise you're talking about - which varies from device to device. Most digital devices put out some, and often significant, ultrasonic noise. This noise is considered to be inaudible, and to interfere with "legitimate measurements"; most high-end test sets - like the AP - deliberately filter out this noise to make their measurements more accurate. This is standard practice and nobody's trying to hide anything. (Just like, when car companies specify how noisy their interior is, they don't count ultrasonic noise that might make your pet bat uncomfortable.) 3) Specifications provided by DAC vendors, especially small boutique companies, can be entirely goofy. Since it's actually rather difficult to measure things like S/N ratios above 100 dB and THDs below 0.01%, and DACs tend to have specs up in that range, it's a common habit for DAC manufacturers to pull the specs for the DAC CHIP they use right off the chip data sheet and provide them as the specs for their unit. This is usually unrealistic... and often downright misleading. For example, a Sabre 9018 has a S/N well over 120 dB - but, odds are, the S/N of your particular DAC device is limited by its analog components, or even by the quality of its clocks and power supply, and not by the DAC chip. (But you'll notice that many low end DACs seem to have exactly the same specs as the ones on the data sheet for the chip they used.  ) 4) Sabre DAC chips have a sort of distinctive sound - which some people like and some don't. People who like it say they sound "very detailed"; people who don't say they sound "etched" or "grainy"; to me it's sort of like the difference between an LED bulb and a halogen bulb. (Sabre DACs have a really good S/N and are really flat, just like other good DAC chips - this is something else to do with their oversampling filters.) And, obviously, tube analog stages have a distinct sound (which we won't discuss here). Also, for the record, no major manufacturer makes a true tube DAC - because it would have to be huge and expensive and wouldn't work very well.... and tubes have pretty much the exactly opposite electrical properties to those you'd want in a DAC's I/V stage (although it has been done).... but you do have quite a few solid state DACs that have a tube buffer stage added onto the end to add some tube coloration to the sound. 5) Once you get past a S/N (or dynamic range) of about 100 dB, it's all sort of moot anyway. Pretty well any good DAC is going to be quieter than any source material you might have. While "good numbers" never hurt, I can't imagine being able to hear a difference between a S/N of 110 dB and one of 120 dB (for reference, a really quiet vinyl album might hit 65 dB or 70 dB, and the S/N on a CD tops out a bit above 85 dB). Likewise, I can't imagine that there's an audible difference between a THD of 0.03% and one of 0.003%. However, all of these little thing just highlight why many people prefer to keep their DAC separate from their player.... so they can pick just the DAC they want, and just the player they want, without having to even hope that their favorite choice for each will happen to coexist in the same unit. |
|
KeithL
Administrator
Posts: 10,274
|
Post by KeithL on May 5, 2016 10:32:25 GMT -5
I did take a look at it - and it looks quite nice - for what it is. It also looks to be VERY well thought out. However, I will have to admit that it is pretty much the antithesis of what I personally look for in a music player. My background is in computers at least as much as it is in music. And, beyond that, I guess I'm what you would call "data centric". When I think about my favorite album, I don't see it in terms of the group who recorded it, and the people in that group; I see it as a bunch of files (songs), in a folder (the album), in a bigger folder (which contains all that group's albums). I have very little interest in seeing the album cover, or hearing news about what the members of that group are doing today, and, if I start to wonder, I just look it up on Google (on the computer that's stays on 24/7 in my living room). And I just see all that other stuff as distractions. But what I do value is universality. I want to be able to download a song, put it on a stick, double click on the file, and have it play... without worrying if it's in the right format, or importing it into my library, or indexing it. And I really don't see it as a hassle to open my file browser, go to the album folder, and double-click on the song I want to listen to - or on a playlist. I only usually play music in my living room, but, if I wanted to play it in other rooms, I would simply share that drive across the network. That's pretty well a summary of the reasons I stopped using the Squeezebox: I'd download a new album, or rip a CD, and then have to go through a process to get it into my library; and, even beyond that, I often found music files that it wouldn't play, or that it insisted on converting to different formats, or that it would seem to index and put in odd places where I couldn't find them. In the end I found it easier to just put all my music on a big drive, in standard FLAC format, in hierarchical folders, and use something like Foobar2000, which plays pretty well every audio format known to science, without converting it, at the proper sample rate. I also use jRiver, but I usually turn off the indexing and just use it to play files by clicking on them. And, since my files are "indexed physically", I can play the same files from the same library without having to specifically set them up for Foobar or jRiver, or whatever... I just use my favorite file manager - which is something called Total Commander. Again, though, I just look at my music as another type of computer file - to be played using the appropriate program... double click on it and it plays. However,as I said at the start, for those who prefer a "rich music environment", and find the features it offers attractive, Roon seems to be very well thought out - and well worth the price (if it lives up to claims). Not that I disagree with you Keith but there are more than one way to skin the flexibility cat, not all of ut has to be done with a full fledged computer. Since you're familiar with the squeezebox infrastructure; Unless you already have, look at Roon. |
|
|
|
Post by yves on May 5, 2016 15:04:34 GMT -5
The notion that a better wideband SNR is still key to advancing fidelity comes from an outdated concept of resolution. While a low widened SNR not necessarily say anything about the more important SNR limited to hearing range or A weighted, a high wide band SNR implies a high a SNR limited to the hearing range as well as A weighted SNR. The inherent noise shaping of Delta Sigma pushes quantization noise out of the audible band (i.e. pushes it out of the <20kHz region into the >20kHz region, where it can be easily filtered). It [Delta Sigma noise shaping] very substantially improves accuracy (Dynamic Range) if heavy upsampling (8× internal upsampling in the ES9018) is used in conjunction with a high order (5th order in the ES9018) Delta Sigma Modulator. In the case of modern Delta Sigma DAC, it typically just means the anti-imaging filter used is so sharp the filter introduces, into the audible band, more severe ringing artifacts. That is, when compared to using a slow roll-off filter that could have prevented imaging artifacts just as adequately as the sharp one. Furthermore, if heavy upsampling is used in conjunction with paralleled DAC (8 mono channels can be paralleled to a single mono channel in the ES9018), purposely leaving some of the supersonic noise unfiltered so that it folds back into the audible band can actually eliminate Differential Nonlinearity (DNL), kind of like how dither removes quantization distortion. |
|
|
|
Post by yves on May 5, 2016 15:47:26 GMT -5
Obviously the 98 dB SNR measurement is wideband. Interestingly, the $4,995 Mytek does not yield higher Dynamic Range than my $1,350 Eastern Electric; the latter brings on-the-fly switching between solid state gain stage and tube based gain stage to the table, and can be upgraded (optionally) with femto clocks and discrete op-amps because the design is modular of course so the buyer gets the additional freedom of choice on whether to break the bank towards scraping that last grain of sand from under the soles of the pair of ES9018 chips or towards optimum balance of maybe drinking more excellent beer instead. I like that kind of freedom a lot. There's nothing obvious about not stating whether a measurement is, wide band, unweighted or A weighted, on the contrary. In the case of modern Delta Sigma DAC expressing a whopping 130 dB Dynamic Range, the fact 98 dB SNR is wideband is clear as a bell unless you are new to Delta Sigma, and, if you are new to Delta Sigma, that means you should start reading up on it, as opposed to trying to lecture. Wrong. Unlike the SNR, Dynamic Range in audio engineering takes distortion into account so please re-read the Wikipedia article before hitting the reply button again. Fair enough, but it [my DAC] uses dual toroidal power transformers and dual ES9018 chips (one for each channel), proper galvanic isolation, ferrite shunts and whatnot. So OF COURSE it is going to be more silent than a grave. So then, don't compare the prices? Instead of comparing them, I just pocketed the money, sat back, and enjoyed like no tomorrow. I wouldn't tell you about my personal finances even if you *were* interested, simply because that's totally none of your business anyway. |
|
|
|
Post by qdtjni on May 5, 2016 16:03:50 GMT -5
.... gets the choice of whether to break the bank towards scraping that last grain of sand from under the soles of the pair of ES9018 chips or towards optimum balance of maybe drinking more excellent beer instead. I like that kind of freedom a lot. I wouldn't tell you about my personal finances even if you *were* interested, simply because that's totally none of your business anyway. Dude, you already did.  |
|
|
|
Post by yves on May 5, 2016 17:12:11 GMT -5
.... gets the choice of whether to break the bank towards scraping that last grain of sand from under the soles of the pair of ES9018 chips or towards optimum balance of maybe drinking more excellent beer instead. I like that kind of freedom a lot. I wouldn't tell you about my personal finances even if you *were* interested, simply because that's totally none of your business anyway. Dude, you already did. I said *the buyer* gets the choice; the breaking the bank remark was just a figure of speech to point out that things from the likes of femto clocks are a wee bit over the top IMO, even though I have no trouble admitting this is still only my own subjective opinion. |
|
|
|
Post by JKCashin on May 5, 2016 23:29:34 GMT -5
So here's a question then. Optical or USB from my Mac Mini to my XMC-1? If you want to play source material with higher sampling rate than 96 kHz and not use DIRAC, use USB. If you want to use DIRAC, either USB or Optical with your player software downsample to 96 kHz. Use the one that find sounding best and/or is more convenient. Thanks! That's the info I was looking for. I forgot about the Dirac limit of 96 kHz |
|
|
|
Post by yves on May 6, 2016 5:29:54 GMT -5
I did take a look at it - and it looks quite nice - for what it is. It also looks to be VERY well thought out. However, I will have to admit that it is pretty much the antithesis of what I personally look for in a music player. My background is in computers at least as much as it is in music. And, beyond that, I guess I'm what you would call "data centric". When I think about my favorite album, I don't see it in terms of the group who recorded it, and the people in that group; I see it as a bunch of files (songs), in a folder (the album), in a bigger folder (which contains all that group's albums). I have very little interest in seeing the album cover, or hearing news about what the members of that group are doing today, and, if I start to wonder, I just look it up on Google (on the computer that's stays on 24/7 in my living room). And I just see all that other stuff as distractions. But what I do value is universality. I want to be able to download a song, put it on a stick, double click on the file, and have it play... without worrying if it's in the right format, or importing it into my library, or indexing it. And I really don't see it as a hassle to open my file browser, go to the album folder, and double-click on the song I want to listen to - or on a playlist. I only usually play music in my living room, but, if I wanted to play it in other rooms, I would simply share that drive across the network. That's pretty well a summary of the reasons I stopped using the Squeezebox: I'd download a new album, or rip a CD, and then have to go through a process to get it into my library; and, even beyond that, I often found music files that it wouldn't play, or that it insisted on converting to different formats, or that it would seem to index and put in odd places where I couldn't find them. In the end I found it easier to just put all my music on a big drive, in standard FLAC format, in hierarchical folders, and use something like Foobar2000, which plays pretty well every audio format known to science, without converting it, at the proper sample rate. I also use jRiver, but I usually turn off the indexing and just use it to play files by clicking on them. And, since my files are "indexed physically", I can play the same files from the same library without having to specifically set them up for Foobar or jRiver, or whatever... I just use my favorite file manager - which is something called Total Commander. Again, though, I just look at my music as another type of computer file - to be played using the appropriate program... double click on it and it plays. However,as I said at the start, for those who prefer a "rich music environment", and find the features it offers attractive, Roon seems to be very well thought out - and well worth the price (if it lives up to claims). Not that I disagree with you Keith but there are more than one way to skin the flexibility cat, not all of ut has to be done with a full fledged computer. Since you're familiar with the squeezebox infrastructure; Unless you already have, look at Roon. I don't have time to look at "rich music environment" because I'm too busy trying to tell people on The Lounge etc. how wrong they usually are! J/K... |
|
|
|
Post by qdtjni on May 6, 2016 11:47:06 GMT -5
If you want to play source material with higher sampling rate than 96 kHz and not use DIRAC, use USB. If you want to use DIRAC, either USB or Optical with your player software downsample to 96 kHz. Use the one that find sounding best and/or is more convenient. Thanks! That's the info I was looking for. I forgot about the Dirac limit of 96 kHz Juts to be clear, Dirac is limited to 48 kHz in XMC-1 and the toslink in recent Macs to 96 kHz. |
|
|
|
Post by yves on May 6, 2016 23:14:05 GMT -5
|
|
)
